3.5 Endosteum dictates the age-related
compressive or tensile principal strains dominate the other. In the following, using a similar approach to section 3.4, (re)modeling probabilities were deter-mined at endosteal and periosteal surfaces for sites mainly under tension and compression, separately.
Sites mainly under compression
In young animals, additional loading resulted in an increase in the proba-bility of formation events with increasing strain levels at both endosteal and periosteal surfaces (Fig. 3.26). In this age group, additional loading resulted in a significant increase in the formation probability at all strain levels (low, medium and high) at the endosteum which reached ∼95% maximum chance.
At the periosteum, the formation probability significantly increased at medium and high strain levels which reached ∼85% maximum chance. The average formation probability was higher (∼10%) at the endosteal surface compared with the periosteal surface at each given strain level in young animals. In the contra-lateral limbs undergoing physiological loading, a similar difference in the average formation probability between endosteal and periosteal surfaces was observed at medium and high strain levels (Fig. 3.26).
In adult animals, additional loading resulted in an increase in the proba-bility of formation events at both endosteal and periosteal surfaces. At both surfaces, formation probability significantly increased at medium and high strain levels (Fig. 3.26). The probability of formation events at the periosteal surfaces were significantly smaller than at the endosteum at medium and high strain levels. Moreover, in response to additional loading, an increase in the formation probability at high strain levels at the periosteum was observed (Fig. 3.26) which could explain the double-peak observed in the formation
probability in figure 3.23.
In elderly animals, additional loading resulted in an increase in the prob-ability of formation events at medium and high strain levels at both bone surfaces (Fig. 3.26). In the both contra-lateral limbs, the formation proba-bility at the endosteum was not significantly higher than at the periosteum (Fig. 3.26).
Figure 3.26: Probability of formation events at each available compressive strain (minimum principal strain) at endosteal (solid line) and periosteal (dashed line) surfaces in physiologically (top) and additionally (bottom) loaded limbs in 10 (n = 5), 26 (n = 10), and 78 (n = 9) week old mice. At each specific strain level, the sum of formation, resorption and quiescence frequencies at each surface (endosteum or periosteum) is 1 (i.e., 100%). (Re)modeling sites were determined following 15 days of the loading experiment. Plots show mean (line) and standard deviation (shaded area along line).
Resorption events in young animals, as mentioned earlier (section 3.4), were too few to be evaluated (at both surfaces and both limb sides). In adult animals, additional loading resulted in a significant decrease in the probability of resorption events at medium and high strain levels at the endosteal surface (compared with physiologically loaded limb; Fig. 3.27). At the endosteal surface, the average resorption probability in the physiologically loaded limbs was ∼50% at all strain levels, this chance was sharply reduced in response to additional loading at medium and high strain levels (Fig. 3.27).
At the periosteal surface, a significant reduction in the resorption probability is response to additional loading could not be observed; however, in both contra-lateral limbs at all strain levels, the probability of resorption events at the periosteum was significantly smaller than at the endosteum.
Figure 3.27: Probability of resorption events at each available compressive strain (minimum principal strain) at endosteal (solid line) and periosteal (dashed line) surfaces in physiologically (top) and additionally (bottom) loaded limbs in 10 (n = 5), 26 (n = 10), and 78 (n = 9) week old mice. At each specific strain level, the sum of formation, resorption and quiescence frequencies at each surface (endosteum or periosteum) is 1 (i.e., 100%). (Re)modeling sites were determined following 15 days of the loading experiment. Plots show mean (line) and standard deviation (shaded area along line).
In elderly animals at the endosteal surface, additional loading resulted in a significant decrease in the probability of resorption events at the medium and high strain levels (Fig. 3.27). However, at medium strain levels at the endosteum, additional loading could only partially and not fully suppress the resorption probability.
In addition to the above-mentioned observations, in the additionally loaded limbs in adult animals, a significant difference was observed between forma-tion and resorpforma-tion probabilities at all strain levels at the endosteum, i.e. at the endosteum a significantly higher resorption probability compared with formation was observed at low strain level and a significantly higher formation probability compared with resorption was observed at medium and high strain levels (Fig. 3.26 and 3.27). In elderly animals, however, only at low strain
level at the endosteum significantly higher resorption probability compared with formation probability was observed (in additionally loaded limbs).
In this sense, a clear bone response, e.g. either formation or resorption, at the medium strain levels could be observed in adult animals at the endos-teum while elderly animals showed similar probability of bone formation and resorption at medium strain levels. In addition, it was observed that at the periosteum in adult and elderly animals, the probability of formation and resorption events at medium strains were not significantly different.
Based on these observations, the dysfunction of elderly animals to elicit a clear formation or resorption response at the medium strain levels (described in section 3.4.3) is essentially an effect which can be seen at the endosteal surfaces and not the periosteal. Considering that these results show the prob-ability of the (re)modeling events in relationship to strain, it is now crucial to know how often formation/resorption occurs at a strain level at each of these two surfaces, i.e. the percentage of the spotted formation/resorption events at each of these two bone surfaces. It could then be revealed which percentage of the mechanoregulation observed in each surface is actually contributing to the responsiveness of bone to loading with age.
For sites mainly under compression, table 3.2 lists the percentage of the sum of the occurrence of each (re)modeling event at the endosteal and pe-riosteal surfaces, i.e. the formation spots are shown in percentage relative to the total spots available at both bone surfaces. Additional data regarding this table is given in Appendix B.1 (Fig. B.1).
In young animals in the additionally loaded limbs, the sum of the total occurrence frequency of formation events at the periosteal surface (20%) was larger compared with the endosteal surface (15.6%), (Table 3.2). Considering that a larger surface area at the periosteum is mainly under compression (36.5%) than endosteum (24%), the relative effect is inverse, i.e. the endosteal surface in young animals show more anabolic response to strains than the periosteum.
In adult and elderly animals in the additionally loaded limbs, the sum of total occurrence frequencies of formation events at both surfaces were comparable (8%: endosteal and 7%: periosteal in adults, and ∼5%: endosteal and periosteal in elderly). The sum of occurrence frequencies of resorption events in both additionally and physiologically loaded limbs were larger at the endosteal surface compared with the periosteum in adult and elderly animals.
This data shows that a pivotal difference in the probability of formation events between young and adult animals occurred at the periosteum and at low strain levels of the endosteal surface. Between adult and elderly, however, a strong reduction in the probability of formation events was observed at the endosteal surfaces.
Table 3.2: Sum of the total (re)modeling occurrences (formation, resorption and quiescence) at compressive strain sites in both limbs. Average values are given in 10 (n = 5), 26 (n = 10), and 78 (n = 9) week old mice.
Mouse Limb (Re)modeling probabilities
Age Side Endosteal Periosteal
Form Resorb Quies. Total Form Resorb Quies. Total 10 L 15.6% <1% 8.5% 24.2% 20.0% <1% 16.5% 36.5%
10 R 6.8% <1% 20.0% 27.0% 3.6% <1% 19.4% 23.6%
26 L 8.0% 5.0% 13.0% 26.0% 7.0% <1% 34.0% 41.0%
26 R <1% 12.4% 16.4% 29.6% 1.1% 1.2% 30.1% 32.6%
78 L 5.4% 7.3% 20% 32.7% 5.1% 2.0% 27.4% 34.5%
78 R 2.5% 13.5% 19.6% 35.6% <1% <1% 32.0% 33.4%
Sites mainly under tension
In the section 3.4, a dissimilar mechanoregulatory behavior was observed at sites which were mainly under tension compared with sites mainly under compression. Specifically, it was shown that in adult animals a strong decrease in the formation probability occurred at high tensile strain levels but not at high compressive strain levels (shown in figure 3.24). In this section, the mechanoregulation of bone at sites mainly under tension at the endosteum and periosteum is presented.
In young animals in response to additional loading, the formation proba-bility at both bone surfaces surface significantly increased compared with the background response (similar to the compression strains; Fig. 3.26 and 3.28).
However, unlike compression sites, formation events at the endosteum in the additionally loaded limb did not reach ∼100% maximum chance at highest strain levels. Perhaps, this effect owed to the smaller tensile strains induced at the endosteum compared with the compressive strain levels at that surface
and to the background response observed in the physiologically loaded limb (Fig. 3.26 and 3.28).
Figure 3.28: Probability of formation events at each available tensile strain (maximum principal strain) at endosteal (solid line) and periosteal (dashed line) surfaces in physiologically (top) and additionally (bottom) loaded limbs in 10 (n = 5), 26 (n = 10), and 78 (n = 9) week old mice. At each specific strain level, the sum of formation, resorption and quiescence frequencies at each surface (endosteum or periosteum) is 1 (i.e., 100%). (Re)modeling sites were determined following 15 days of the loading experiment. Plots show mean (line) and standard deviation (shaded area along line).
In adult animals, additional loading resulted in a significant increase in the probability of formation events at medium and high strain levels at the endosteum (compared with the physiologically loaded limb). Interestingly, the formation probability in adult animals in response to additional loading at the endosteal surface was comparable with young animals (inter-limb differences, Fig. 3.28). At the periosteal surface, a response to additional loading was not observed in this age group (Fig. 3.28). In elderly animals additional loading lead to an increase in the formation probability at all strain levels at the periosteum; however, no specificity between the local strain level and the formation probability could be observed (i.e. similar formation probability at
all strain levels).
Additional loading resulted in a decrease in the resorption probability above a certain strain level at endosteal surfaces in adult animals (Fig. 3.29).
In elderly animals, a similar effect at the endosteal surface could not be observed (Fig. 3.29). Likewise the anabolic responses, a major dysfunction in the mechanoregulation of the resorption probability between adult and elderly animals occurred at the endosteal surfaces.
Figure 3.29: Probability of resorption events at each available tensile strain (maximum principal strain) at endosteal (solid line) and periosteal (dashed line) surfaces in physiologically (top) and additionally (bottom) loaded limbs in 10 (n = 5), 26 (n = 10), and 78 (n = 9) week old mice. At each specific strain level, the sum of formation, resorption and quiescence frequencies at each surface (endosteum or periosteum) is 1 (i.e., 100%). (Re)modeling sites were determined following 15 days of the loading experiment. Plots show mean (line) and standard deviation (shaded area along line).
The amount of endosteal and periosteal bone surfaces undergoing for-mation/resorption activities is an important factor to explain the above-mentioned observations. For sites mainly under tension, table 3.3 lists the
sum of the occurrence of each given (re)modeling event at the endosteal and periosteal surfaces, e.g. the formation spots are shown in percentage relative to the total spots available at the bone surface. Aditional information regard-ing this table is given in figure B.2 in the supplemental data (Appendix B.1).
In young animals, the total sum of the formation occurrence frequencies in response to additional loading were larger at the periosteum (12%) than at the endosteum (7%) (Table 3.3). But relative to the larger surface under tension at the periosteum (25%) than endosteum (14%), it can be concluded that the formation occurrence frequencies at both surfaces in response to additional loading were comparable. Negligible portions of resorption events occurred at the periosteum compared with endosteum under tension in both adult and elderly animals (Table 3.3). These data reconfirm the observation that endosteal surfaces are in charge for a dysfunction in the mechanical response with age.
Table 3.3: Sum of the total (re)modeling occurrences (formation, resorption and quiescence) at tensile strain sites in both limbs. Average values are given in 10 (n = 5), 26 (n = 10), and 78 (n = 9) week old mice.
Mouse Limb (Re)modeling probabilities
Age Side Endosteal Periosteal
Form Resorb Quies. Total Form Resorb Quies. Total
10 L 7.0% <1% 5.9% 13.0% 12.0% <1% 12.5 24%
10 R 3.6% <1% 11.6% 15.6% 4.5% <1% 28.0% 32.5
26 L 3.0% 1.5% 5.6% 10% 1.7% <1% 21.7% 23.4%
26 R 1.5% 4.2% 8.5% 14.2% 1.2% 1.2% 21.2% 23.6%
78 L 1.5% 5.1% 6.8% 13.0% 1.9% 1.1% 16.6% 19.6%
78 R 1.1% 4.0% 5.4% 10.5% <1% 1.3% 17.0% 18.5%
Based on these results, it can be concluded that with age a separation between strain levels which lead to bone formation or resorption responses to mechanical loading becomes unclear and that this dysfunction mainly occurs at the endosteal surface. Up to this point, the relationship presented between (re)modeling sites and local strain magnitudes were determined after 15 days of the loading experiment. Thus, it remains to be determined how this dysfunction develops over the course of time in animals of different ages.
A specific question addressed is whether elderly animals respond to loading
in a de-controlled manner at different time points during the experiment. To achieve that, in the next section the probability of bone formation/resorption events is presented at specific strain levels at day 5, 10 and compared with day 15 of the loading experiment.